As Washington weighs how to respond to advances in Chinese AI, the industry is urging policymakers against sweeping restrictions on open-weight models — a debate with major consequences for open-source generative AI and synthetic media tools.
AI chip startup Etched has reached a $10.3B valuation from big-name investors, betting its transformer-specialized Sohu ASIC can outrun Nvidia GPUs on inference — with major implications for AI video and synthetic media generation costs.
BOXX Insurance has added affirmative AI and deepfake coverage to its Cyberboxx Business policy, signaling that synthetic media fraud is now a mainstream, insurable enterprise risk.
Substack is rolling out a transparency tool that surfaces which newsletters lean on AI to generate content, adding a new layer of disclosure to the synthetic-text era and raising fresh questions about detection reliability.
New research tackles the 'hidden puppet master' problem: predicting how manipulative LLM dialogues change human beliefs. The work offers a framework for detecting persuasion and covert influence in conversational AI systems.
Synthesia, the AI avatar video platform, is moving beyond pre-rendered training videos into interactive live coaching powered by conversational AI avatars, marking a major shift for synthetic media in enterprise learning.
A new arXiv position paper argues that mainstream AI/ML deepfake research is fundamentally misaligned with the realities of AI-generated non-consensual intimate imagery, leaving detection and mitigation efforts poorly matched to actual victim harm.
District 9 director Neill Blomkamp's new AI-generated short 'Nightborne' relies on ByteDance's Seedance video model, reigniting debate over whether generative AI filmmaking is art or slop.
Substack is rolling out an AI-writing detector powered by Pangram to help readers identify newsletters generated by machines rather than humans, marking a new front in the platform authenticity battle.
Music streamer Deezer now reports that more than half of all songs uploaded daily are AI-generated, a striking data point on how synthetic audio is flooding streaming platforms and what it means for authenticity and detection.
TikTok is piloting AI-powered deepfake detection tools designed to help creators identify and protect against unauthorized synthetic replicas of their likeness, marking a major platform-level push into digital authenticity.
Team DACTYL's PAN 2026 submission combines Bayesian data mixing and empirical cross-risk minimization to build more robust AI-generated text detectors, tackling generalization across unseen generators and domains.